Yu-Jie Zhou1, Gui-Qi Zhu2,3, Xiao-Fan Lu4, Kenneth I Zheng5, Qi-Wen Wang1, Jin-Nan Chen1, Qing-Wei Zhang1, Fang-Rong Yan6, Xiao-Bo Li7. 1. Division of Gastroenterology and Hepatology, Key Laboratory of Gastroenterology and Hepatology, Ministry of Health, Shanghai Institute of Digestive Disease, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, 145 Middle Shandong Road, Shanghai, 200001, China. 2. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China. 3. State Key Laboratory of Genetic Engineering, Fudan University, Shanghai, China. 4. Research Centre of Biostatistics and Computational Pharmacy, China Pharmaceutical University, Nanjing, 210009, China. 5. Department of Hepatology, The First Affiliated Hospital of Wenzhou Medical University, Wenzhou, China. 6. Research Centre of Biostatistics and Computational Pharmacy, China Pharmaceutical University, Nanjing, 210009, China. f.r.yan@163.com. 7. Division of Gastroenterology and Hepatology, Key Laboratory of Gastroenterology and Hepatology, Ministry of Health, Shanghai Institute of Digestive Disease, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, 145 Middle Shandong Road, Shanghai, 200001, China. lxb_1969@163.com.
Abstract
BACKGROUND: Immunotherapy could trigger durable response in advanced gastric cancer, but it only benefits a minority of patients. We aimed to propose a robust molecular classification of gastric cancer microenvironment to identify ideal candidates for tailoring effective immunotherapy. METHODS: A training cohort of 375 gastric cancer samples with RNA sequencing data was analysed. We virtually microdissected tumour, stromal, and immune cell gene expression patterns employing a non-negative matrix factorization algorithm. These expression patterns were annotated using immune- and stromal-related gene signatures. Validation of immunogenomic classification was performed across six microarray datasets of 1406 samples. RESULTS: We found approximately half of gastric cancer samples to have higher immune cell infiltrates, PD-L1 expression, markers of cytolytic activity, and fewer copy number aberrations (all P < 0.05). We termed this group of tumours the Immune Class, which incorporated two components, namely Immune Activation and Immunosuppressive Subtype, according to immunosuppressive or activated microenvironment. Immune Activation Subtype was associated with improved survival in multivariate survival analysis and shared similar genomic characteristics with responders of anti-PD-1 therapy. Immunosuppressive Subtype featured high immune infiltration, stromal enrichment, and transforming growth factor (TGF)-β signalling pathway activation and correlated with non-responsiveness signature of checkpoint blockade therapy, which might be suitable for anti-PD-L1 and anti-TGF-β combined therapy. CONCLUSIONS: We proposed and independently validated three reproducible immune molecular subtypes of gastric cancer, which may provide implications for patient selection of immunotherapy.
BACKGROUND: Immunotherapy could trigger durable response in advanced gastric cancer, but it only benefits a minority of patients. We aimed to propose a robust molecular classification of gastric cancer microenvironment to identify ideal candidates for tailoring effective immunotherapy. METHODS: A training cohort of 375 gastric cancer samples with RNA sequencing data was analysed. We virtually microdissected tumour, stromal, and immune cell gene expression patterns employing a non-negative matrix factorization algorithm. These expression patterns were annotated using immune- and stromal-related gene signatures. Validation of immunogenomic classification was performed across six microarray datasets of 1406 samples. RESULTS: We found approximately half of gastric cancer samples to have higher immune cell infiltrates, PD-L1 expression, markers of cytolytic activity, and fewer copy number aberrations (all P < 0.05). We termed this group of tumours the Immune Class, which incorporated two components, namely Immune Activation and Immunosuppressive Subtype, according to immunosuppressive or activated microenvironment. Immune Activation Subtype was associated with improved survival in multivariate survival analysis and shared similar genomic characteristics with responders of anti-PD-1 therapy. Immunosuppressive Subtype featured high immune infiltration, stromal enrichment, and transforming growth factor (TGF)-β signalling pathway activation and correlated with non-responsiveness signature of checkpoint blockade therapy, which might be suitable for anti-PD-L1 and anti-TGF-β combined therapy. CONCLUSIONS: We proposed and independently validated three reproducible immune molecular subtypes of gastric cancer, which may provide implications for patient selection of immunotherapy.
Authors: Ana Bermúdez; Isabel Arranz-Salas; Silvia Mercado; Juan A López-Villodres; Virginia González; Francisca Ríus; María V Ortega; Carmen Alba; Isabel Hierro; Diego Bermúdez Journal: Diagnostics (Basel) Date: 2021-05-25